Abstract: Veterans are vulnerable to develop complex forms of addictive disorders characterized by the presence of coexistent psychiatric conditions, particularly post-traumatic stress disorder (PTSD) and mood disturbances. Furthermore, the high frequency of traumatic brain injury (TBI) among veterans convey an additional risk to develop refractory forms of substance use disorders probably related to the traumatic disruption of prefrontal circuits mediating reward attribution, impulse control, and emotional regulation. Alcohol use disorders (AUD) are frequently observed among veterans and exert a detrimental effect on community reintegration during the post-deployment period. Acknowledging the significance of AUD, the VA is devoting extraordinary efforts to develop adequate treatment strategies. For instance, the Quad Cities VA Intensive Outpatient Program (IOP) will offer a rehabilitation alternative to the growing number of veterans with AUD in Iowa, Illinois and Northern Missouri. The implementation of this program gives a unique opportunity to evaluate treatment outcomes and to identify specific factors that moderate treatment response. In the present study we will examine treatment outcomes of an IOP administered during a 12-week period. All patients will receive an empirically validated psychosocial intervention and appropriate treatment of coexistent psychiatric conditions. With regard to pharmacological treatment of AUD, we plan to randomize patients to receive either sodium valproate or naltrexone, a standard of care active comparator. The objectives of this study are to determine: 1) whether the presence of prefrontal brain damage will be associated with poorer response to the IOP, 2) if valproate is more effective than naltrexone to treat AUD in this group of patients with severe psychiatric comorbidity and TBI exposure, and 3) if completion of the IOP will be associated with good psychosocial outcomes at 6 months follow-up. This is a double-blind active-controlled randomized clinical trial. Patients will be evaluated at baseline, followed weekly for 12 weeks and evaluated at 6 months. Our primary outcome variable will be time to relapse to heavy drinking. Secondary outcome measures will include proportion of heavy drinking days, severity of PTSD symptoms measured by the Clinician-Administered PTSD Scale, severity of depressive symptoms measured by the Hamilton Depression Rating Scale, and psychosocial outcome measured by the Community Integration Questionnaire. Other assessments will include baseline structural magnetic resonance imaging, diffusion tensor imaging, and a neuropsychological battery emphasizing the assessment of memory and executive functions.